A Closer Look at Engineered Stream Structures for Habitat Restoration

A Closer Look at Engineered Stream Structures for Habitat Restoration


I’ve talked in previous articles about engineered structures and how important they are to the restoration of salmon and steelhead runs. Each engineered structure type is designed to fulfill a specific purpose.  Many of them have multiple benefits.  In this article I’d like to go into a little more detail about specific engineered structures that we have developed through trial and error.

The current favorite structure type is the engineered log jam.  They are used mostly in the larger rivers and can span 45 to 85 feet across the channel or streambank depending on their intended function.  They are excavated up to 20 feet deep into the stream bottom and back 30 feet or more downstream or toward the river bar if used to split a channel.   Rootwads are placed in a configuration across the excavation and used as pilings to anchor the structure.  Logs of various lengths and diameters are positioned and cabled to the rootwads in a pre-designed fashion to add support.  Smaller logs, branches, stumps and the excavated bedload are placed on top of the cross logs to add mass.

This is a very practical structure to build because rootwads and logs are salvaged from land clearing sites to use in the log jams.  This woody debris is more plentiful and easier to transport and acquire than the large whole trees 36-60 inches in diameter and 50 to 70 feet that would be used to produce the same outcome.  Another widely used structure type is the placing of rootwads for bank protection at eroding banks and to provide hiding cover, and to deflect the streamflow.  A new technique coming into use is falling whole trees into the river.  Some are cabled in-place or left to move down the channel to form natural log jams or to form a resting pool. 

The majority of the 2500 structures I placed in my projects were log weirs (also called sill logs), log or rootwad deflectors, digger logs and cover structures.  Log weirs are used to form a rearing pool above the weir log and a plunge pool below.  Some weirs were constructed in a V pointing up or downstream and joined together in the middle of the channel and angled 25 degrees into the stream bottom.  The upstream facing V structures performed best, especially within deposition reaches.  Other weir functions regardless of configuration are to trap and hold bedload consisting of spawning sized gravels, cobble, and small boulders to cover areas scoured down to bedrock, and to create a change in the stream gradient.  Log and rootwad deflectors are positioned 22 or 45 and degrees downstream and angled into the channel.  They can be placed 90 degrees from the streambank and angled 15 to 20 degrees into the stream bottom.  These structures deflect water away from the streambank, forms holding water for migrating fry and adult fish and additional useful hiding cover in the rootmass. 

A digger log is constructed the same as the rootwad and log deflector, only it is angled only 10 to 15 degrees into the streambed.  The digger log is positioned so the streamflow passes over the log as the water level increases, creating holding water.  I would modify the weirs, deflectors and rootwads by adding additional logs or rootwads across them angled parallel to the streambank or 20 to 45 degrees downstream to provide more mass and hiding cover.  All structures need to have several feet of the structure excavated well into the stream bank and boulders placed as rip rap (armoring) to prevent exposing the end of the structure to bank erosion. 

The placement of engineered structures varies depending on many factors such as the type of bedload existing or desired, their position and configuration in the stream channel, and the desired function.  For example a weir may be placed to change stream gradient or trap spawning gravels in one reach (section) and to form a plunge pool for fish to rest and hide from predators in another nearby reach.   

Most of the streambanks where I installed structures were made up of a mixture of sand/silt, cobble and small boulders.  The riparian area (the area a given distance upslope from the active channel) typically consisted of second growth or pole sized conifers and red alder.  These streambank conditions are very prone to erosion especially during high water and flood events.   These conditions result in a wider and shallower channel that could cause the channel to ‘braid” into several shallow channels.  When the stream flow subsides, fry often become stranded in these shallow channels and die from dehydration. 

A healthy stream channel has three characteristics that salmon utilize to complete their life cycle.  They are pools, riffles, and glides.  Pools are needed to provide resting areas and to hold water.  Fish generally need a ratio of 60% pools and 40% riffles (swift areas of fast water).  Riffles are swift flowing sections of the stream channel where spawning gravels, small, or large boulders congregate, especially in deposition areas.  Glides are shallow areas a few inches or feet deep where fish can rest and spawn if there is sufficient flow and the bedload consists of proper sized spawning gravels.  Ideally there would be natural woody debris or engineered cover structures in each channel characteristic to provide protection from predators, including humans.  It has been my experience that most streams fall short of enough woody debris in the channel in a configuration to provide these important characteristics.  Enter the era of the development and use of engineered structures. 

As I gained experience and knowledge, I learned which structure configuration worked in a given reach keeping in mind that the stream condition was always changing and that if the wrong structure type was used it could have an adverse result than intended.  My most often used engineered structures were angled and V weirs, deflectors and cover structures.  Most were constructed using logs 10 to 40 feet long depending on the size of the channel. 

Lessons learned:

•             Weirs placed in low gradient deposition reaches eventually fill both the plunge and rearing pool with bedload and often re-direct the stream flow. 

•             Placing weirs too close together didn’t allow the channel enough distance to adjust.

•             Positioning a log or rootwad deflector angled 45 degrees upstream from each streambank with ends 1/3 to ½ the channel width apart allowed bedload to move downstream and produced a deep pool.  The closer the ends were to each other, the deeper a pool would develop.  This is the best engineered structure configuration to use in deposition areas, instead of a weir or sill log structure type.

•             In some reaches, I was placing structures to close together.  The “rough rule of thumb” is to place structures a minimum of 1 ½ times the stream width apart. 

•             Do not place a log weir at the bend of a river because it that location it will reduce the stream velocity and stop the bedload from moving downstream where it may be better utilized. 


This structure along with the boulder is providing a resting pool and hiding cover for fry and spawning fish and redirecting the streamflow toward the center of the channel.  Above the rootwad is an angled log weir that is collecting bed load, changing the stream gradient and forming a small, shallow plunge pool.  In hindsight, I would have placed deflectors using logs and rootwads in the configuration of small log jams on each end of the weir to form a deeper plunge pool.  In a future article, I will describe how I would make that structure modification.

The brief description of the structure types, their installation, function, placement, size of the logs and rootwads used are meant only as a guide.  Due to the wide variation of stream channel sizes and characteristics, the project designer will have to select each structure location and size and structure type will produce the desired result.  This takes many years of experience by doing and monitoring to determine if the desired results were met.

The goal of restoring habitat and fish runs as you can imagine is something I’m very proud of and consider myself fortunate to have participated in. I hope the information I’ve shared will inspire readers to get involved in stream restoration, and have a better understanding of what goes into recovering our most precious resource, the habitat for fish to spawn in.


Article by Barry Olson

Back to blog